Immunocytochemical localization of the neuron-specific form of the c- src gene product, pp60c-src(+), in rat brain

Neurons express high levels of a variant form of the c-src gene product, denoted pp60c-src(+), which contains a 6 amino acid insert in the amino-terminal half of the c-src protein. We have determined the localization of pp60c-src(+) in neurons using an affinity-purified anti- peptide antibody, referred to as affi-SB12, that exclusively recognizes this neuron-specific form of the c-src gene product. Using affi-SB12, we examined the distribution of pp60c-src(+) by immunoperoxidase staining of sections through adult rat brains, pp60c-src(+) was widely distributed in rat brain and appeared to be differentially expressed in subpopulations of neurons. The majority of immunoreactive neurons was found in the mesencephalon, cerebellum, pons, and medulla. Telencephalic structures that contained substantial populations of pp60c-src(+)-immunoreactive neurons included layer V of the cerebral cortex and the ventral pallidum. Within individual neurons, pp60c- src(+) immunoreactivity was localized to the cell soma and dendritic processes, while labeling of axons and nerve terminals (puncta) was not as readily detected. Dense accumulations of immunoreactive axons were rare, being most prominent in portions of the inferior and superior olive, and in the spinal trigeminal nucleus. While the regional distribution of pp60c-src(+) immunoreactivity does not correlate with any specific neuronal cell type or first messenger system, this unique pattern of expression of pp60c-src(+) suggests the existence of a previously uncharacterized functional organization within the brain. Furthermore, the localization of this neuron-specific tyrosine kinase in functionally important areas of the nerve cell, namely, dendritic processes, axons, and nerve terminals, suggests that pp60c-src(+) may regulate pleiotropic functions in specific classes of neurons in the adult central nervous system.

Neurons express high levels of a variant form of the c-src gene product, denoted pp60c-or~+), which contains a 6 amino acid insert in the amino-terminal half of the C-SK protein. We have determined the localization of pp60C-0@+) in neurons using an affinity-purified anti-peptide antibody, referred to as affiSB12, that exclusively recognizes this neuron-specific form of the C-SC gene product.
Using affiSB12, we examined the distribution of pp60C-0r=(+) by immunoperoxidase staining of sections through adult rat brains. pp6oC-=*+I was widely distributed in rat brain and appeared to be differentially expressed in subpopulations of neurons. The majority of immunoreactive neurons was found in the mesencephalon, cerebellum, pons, and medulla. Telencephalic structures that contained substantial populations of pp60c-s~+)-immunoreactive neurons included layer V of the cerebral cortex and the ventral pallidum. Within individual neurons, pp60c-=*+) immunoreactivity was localized to the cell soma and dendritic processes, while labeling of axons and nerve terminals (puncta) was not as readily detected. Dense accumulations of immunoreactive axons were rare, being most prominent in portions of the inferior and superior olive, and in the spinal trigeminal nucleus. While the regional distribution of p~60-~@+) immunoreactivity does not correlate with any specific neuronal cell type or first messenger system, this unique pattern of expression of p~60~-=~*+) suggests the existence of a previously uncharacterized functional organization within the brain. Furthermore, the localization of this neuron-specific tyrosine kinase in functionally important areas of the nerve cell, namely, dendritic processes, axons, and nerve terminals, suggests that pp60c-gr*+) may regulate pleiotropic functions in specific classes of neurons in the adult central nervous system.
The proto-oncogene c-src is homologous to the retroviral transforming gene v-src of Rous sarcoma virus (for reviews, see Bishop, 1983;Golden and Brugge, 1988;Cooper, 1989). This cellular gene is highly conserved throughout evolution and has been found in such widely divergent species as human Anderson et al., 1985;Sorge et al., 1985;Bolen et al., 1987), chicken (Stehelin et al., 1976;Takeya and Hanafusa, 1983) fish (Bamekow et al., 1982) fruit fly (Shilo and Weinberg, 1981;Hoffman-Falk et al., 1983;Simon et al., 1983;Gregory et al., 1987), and sponge . The c-src gene encodes a 60 kDa phosphoprotein, p~6@:-~~~, which functions as a tyrosine-specific protein kinase (for reviews, see Hunter and Cooper, 1985;Golden and Brugge, 1988;Cooper, 1989). To date the physiological role of pp60C-src has not been identified. However, several lines of evidence suggest that the c-src gene product may serve a specialized function in neurons. High levels of pp6@-src have been detected in brain and in other neural tissues of vertebrates (Cotton and Brugge, 1983;Gessler and Bamekow, 1984;Levy et al., 1984;Bolen et al., 1985;Fults et al., 1985;Shores et al., 1987;Cartwright et al., 1988) and in Drosophila (Simon et al., 1985). The expression of c-src in neural tissues appears to be developmentally regulated as pp6@-src is initially expressed at the onset of neuronal differentiation and then is maintained at high levels in postmitotic, fully differentiated neurons in the adult central nervous system Sorge et al., 1984;Fults et al., 1985;Simon et al., 1985;Lynch et al., 1986;Maness, 1986;Maness et al., 1986;Vardimon et al., 1986;Cartwright et al., 1987Cartwright et al., , 1988Gregory et al., 1987;LeBeau et al., 1987;Mellstrom et al., 1987;Wiestler and Walter, 1988). A unique form of the c-src gene product has since been identified in neurons (Brugge et al., 1985). This protein, designated pp60 csrd+), is expressed at high levels in neurons and is structurally and enzymatically distinct from the p~6oC-~~~ molecule expressed in non-neuronal cells (Brugge et al., 1985(Brugge et al., , 1987a. This neuron-specific form of the C-SK gene product, which contains a 6 amino acid insert in the aminoterminal half of the c-src protein, is encoded by a unique c-src mRNA. This alternately spliced c-src transcript found in neurons contains an 18 base-pair insertion which was mapped directly to the intron between exons 3 and 4 of the c-src gene (Levy et al., 1987;Martinez et al., 1987;Raulf et al., 1989a). This 18 base-pair mini-exon has been detected in chick (Levy et al., 1987), mouse (Martinez et al., 1987) human (Pyper and Bolen, 1989), and fish (Raulf et al., 1989a, b), but is absent in Hydra (Raulf et al., 1989a). The 6 amino acids encoded by the neuron-specific insert are completely conserved in chicken (Levy et al., 1987) mouse (Martinez et al., 1987) and human (Pyper Sugrue et al. l Localization of pp60C-s,*+' i n Rat Brain and , whereas only 50% of these amino acids are conserved in fish (Raulf et al., 1989a). The specific activity of this variant form of c-src protein is elevated compared to that of pp60=, as measured by tyrosine phosphorylation of the src substrate, ~36, in vivo and by anti-phosphotyrosine immunoblots (Brugge et al., 1985;Cartwrigbt et al., 1987;Levy and Brugge, 1989).
Several lines of evidence indicate that p~6@-~"+) is expressed specifically by neuronal cells. pp60C+'d+) is the major form of c-src protein expressed in primary neuron cultures, and it is not detected in cultured astrocytes or fibroblasts or in any nonneural tissues (Cotton and Brugge, 1983;Brugge et al., 1985;Pyper and Bolen, 1989). In addition, embryonal carcinoma cells that have been induced to differentiate into neuron-like cells by the addition of retinoic acid express high levels of p~6@-~"+) (Lynch et al., 1986). Similarly, this variant form of the c-src protein is expressed by primary neurons from rat embryo striaturn that differentiate in culture (Cartwright et al., 1987). Several different human neuroblastoma cell lines express p~6oC-~~d+), whereas this unique form of the c-src gene product has not been detected in any glioblastoma cell lines (Bolen et al., 1985;Mellstrijm et al., 1987;Yang and Walter, 1988). In addition to these in vitro analyses, there is evidence from in vivo studies that p~6@-~~@+) is neuron-specific.
Biochemical studies using mutant mice that display progressive degeneration of specific types of neurons indicate that there is a direct correlation between loss of neurons and a decrease in the levels of p~6o"-'~ti+) (Brugge et al., 1987b). Moreover, neurochemical lesions ofthe rat caudateputamen result in substantial losses of p~6@-~~d'+) in both the striatum and the substantia nigra, which is a major target nucleus for striatal efferents (Walaas et al., 1988). Together, these results provide further evidence that pp60 c II~(+) -is expressed by neurons in vivo.
The expression of a neuron-specific form of the c-src gene product suggests that pp60 C srti+) may play an important role in neuronal cell function. In an attempt to obtain further information on the functional significance of this tyrosine kinase in neurons, we have localized pp60 -c prd+) in rat brain using an affinity-purified polyclonal anti-peptide antibody that exclusively recognizes only the variant form of the c-src protein without cross-reacting with p~6Cl:-~"~.

Materials and Methods
Antibodies. The peptide Asn-Asn-Thr-Arg-Lys-Val-Asp-Val-Arg-Glu-Gly-Asp, which contains the 6 amino acid insert of p~60+~'<+) (underlined), was designed as an antigen to produce a specific antibody to the insert region. Synthesis was performed by condensation of symmetric anhydrides of N-cu-t-Boc-protected amino acids (Barany and Menifield, 1979) on an Applied Biosystems 430A instrument using dimethyl formamide as the coupling solvent. Double coupling cycles were used for Asn and Arg, which were coupled as activated esters using 1 -hydroxybenzotriazole. The side chains were protected as Tosyl-Arg, Asp-P-O-Benzyl, Thr-0-Benzyl, Glu-7-0-Benzyl, and Lys-N-e-chloro-benzyloxycarbonyl (all obtained from Applied Biosystems). The peptide was synthesized on a p-methyl-benzhydrylamine-derivatized, divinylbenzene-cross-linked, polystyrene resin, and the peptide cleaved from the support with liquid HF containing 5% (vol/vol) each of anisole and methylsulfide (Aldrich). The crude peptide was precipitated with diethyl ether, lyophilized, and redissolved in 10% (vol/vol) acetic acid. The peptide was desalted by gel filtration on a column (5 x 100 cm) of BioGel P-2 (BioRad) a&amide resin and lyophilized. Final purification of the peptide was done by reverse phase HPLC on an octadecylsilanyl silica column (Waters). Amino acid analysis and sequence analysis by automated Edman degradation was used to verify the structure of the final product. The peptide was conjugated to Keyhole limpet hemocyanin in the presence ofglutaraldehyde (Kagan and Glick, 1979). The conjugate was then used to immunize rabbits. The antisera obtained from these rabbits were screened for immunoreactivity with pp6@-src and p~6@-*~<+) in immunoblot and immunoprecipitation assays. One pp6@-s"+)-specific antiserum, referred to as SB12, was affinity-purified and then used in all localization studies of p~6@-~~<+' described in this report. Monoclonal antibody 327 was prepared from mice immunized with p~6@-*~~ expressed in Escherichia coli (Lipsich et al., 1983).
Ajinity pur#ication of pp60+src~+)-specific antibodies. The pp60'-srd+)specific antibodies were affinitv-purified from the SB I2 antiserum using a-pp6@<+) affinity column. The-c-src( +) protein that was used As ligand in preparing the affinity column was expressed in bacteria. The c-src(+) cDNA clone (Levy and Brugge, 1989) was inserted into an expression plasmid containing the T7 promoter (Rosenberg et al., 1987;M. Sugrue and J. Brugge, unpublished results). This construct was then used to transform the DE3 strain of E. coli that carries an integrated copy of the T7 polymerase gene (Rosenberg et al., 1987). pp6@:-$,d+) was isolated from these cells and coupled to CNBr-activated Sepharose 4B beads (Pharmacia, Piscataway, NJ). After application of antiserum to the column, the column was washed with 50 mM glycine, pH 2.3, and bound antibody was eluted with 50% ethylene glycol, pH 10.5.
Immunoprecipitation.CEFs were labeled for 4 hr with 1 mCi/ml 32Porthophosphate in phosphate-free Dulbecco's modified medium. The cells were lvsed in RIPA buffer (1% sodium deoxvcholate. 1% Triton X-100, 15s mM NaCl, 5 mM EDTA, 10 mM Tris hydrochloride [pH 7.21, 0.1% SDS). Following clarification, the protein concentration was determined using the method of Lowry et al. (195 1) and the lysates were adjusted to equal protein concentration. The clarified lysates were incubated with antibody and the immunoprecipitates were adsorbed to formalin-fixed Staphylococcus aureus (Pansorbin; Calbiochem-Behring San Diego, CA) as described previously (Levy and Brugge, 1989). The samples were subjected to electrophoresis on 7.5% SDS-polyacrylamide gels (Laemmli, 1970). The dried gels were exposed to X-Omat (Kodak) film using Lightning Plus (DuPont) intensifying screens at -70°C.
Indirect immunojluorescent cell staining. CEFs expressing high levels of either p~6@;-~,~ or pp6@d+) were seeded onto acid-etched 12 mm glass coverslips at 37°C. All subsequent steps were carried out at room temperature (RT). The cells were fixed using freshly prepared 2% formaldehyde in PBS for 20 min and then washed with 0.1 M glycine in PBS for 5 min. The coverslins were washed with PBS containing 0.1% BSA. 0.02% NaN,, and 0.2%Triton X-100 (PBSAT) for 15 min. Fixed cover: slips were either stored in PBSAT at 4°C or used immediately. Prior to the application of primary antibody, the cells were permeabilized with PBS containing 1% NP40,O. 1% BSA for 10 min, and then blocked with PBS containing 5% BSA for 30 min. A 1: 100 dilution of affi-SB 12 was applied to each coverslip and incubated for 1 hr. Then the coverslips were washed in PBSAT for 30 min. Biotinylated anti-rabbit IgG F(ab), (Amersham) was diluted 1: 100 using PBSAT and was applied to each coverslip for 1 hr. After washing with PBSAT for 30 min, FITC Streptavidin (Vector Laboratories, Burlingame, CA) was diluted 1:200 and then added for 1 hr. Finally, the coverslips were washed with PBS and then mounted onto microscope slides. Cells were observed using a Nikon microscope equipped with epifluorescent illumination.
Immunoperoxidase staining of jixed tissue sections. Male Sprague-Dawley rats (150-250 gm) were anesthetized with chloral hydrate (400 mg/kg, i.p.) and perfused transcardially with 200 ml of saline followed by 500 ml of freshly prepared 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.5. Brains were removed, post-fixed for 1 hr at 4"C, and then allowed to sink in 20% phosphate-buffered sucrose overnight at 4°C. Next, 50 Frn-thick coronal or sagittal sections were cut using a vibratome. The sections were rinsed in 10 mM PBS, pH 7.5, and then blocked in 10% normal goat serum (Vector) for 10 min. Sections were then incubated with a 1:5 dilution of affinity-purified SB12 antibody, affi-SB12, overnight at RT. A modification of the avidin-biotin technique of Hsu et al. (198 1) was used to localize p~6@-~~6+) immunoreactivity. After being washed in PBS, the sections were first incubated with a 1:200 dilution of biotinylated goat anti-rabbit IgG (Vector) for 2 hr at RT and then with the ABC complex (Vector) for 2 hr at RT. The peroxidase reaction product was visualized using 0.04% (wt/vol), 3,3diaminobenzidine (DAB, Sigma) and 0.012% (vol/vol) H,O, in 0.1 M phosphate buffer, pH 7.4. Sections were mounted on gelatin-coated slides and allowed to air-dry overnight at RT. Then the sections were dehydrated in an ascending series of alcohols, cleared in xylene, and a coverslip was applied using Permonnt (Fisher). A series of sections adjacent to those processed for immunocytochemistry was stained with cresyl violet for use in the identification of nuclear boundaries. Light microscopy. The schematic diagrams shown in Figure 3 were drawn from the cresyl violet-stained sections described above, and the pp60'+6+)-immunoreactive neurons from the adjacent immunocytochemically processed sections were then plotted onto the drawings using a camera lucida attached to a Leitz microscope. Each dot on these diagrams represents approximately 5 cells. The distribution was taken from a single case and is representative of all the animals used in the study.
Immunolocalization of pp60"src(+) in chick embryo jbroblasts In order to prepare the SB 12 antiserum for use in immunolocalization studies, we affinity-purified the pp60e-srN+)-specific antibodies using an affinity column containing bound pp60c-srd+) and used affi-SB 12 in immunofluorescent cell staining of CEF overexpressor cells, CEFlc-src( +) and CEFlc-src. A positive signal of immunofluorescence was observed in the CEFlc-src( +) cells alone. A distinct pattern of immunofluorescence was observed in these fibroblasts expressing high levels of pp60e+d+). As shown in Figure 2, staining was observed throughout the cytoplasm with enhanced staining in the perinuclear region of the cell. No immunofluorescence was detected when CEFlc-src cells were incubated with affi-SB 12 (data not shown). A similar absence of immunofluorescence was observed when affi-SB12 was preincubated with the c-src(+) peptide antigen (data not shown). These results indicate that affi-SB12 is specific for p~60E-~'4+) and does not cross-react with pp6@-".
Distribution of pp60c-src(+) immunoreactivity in rat brain In most areas where perikarya could be visualized, pp60C-sr~+) immunoreactivity was localized both in the cell body and in the dendritic processes of neurons. In contrast, staining of axons and axon terminals was not as readily detected, being most prominent in brain-stem structures. In the description of the distribution of pp60C-srd+) immunoreactivity below, presumed axons or axon terminals are referred to as puncta or varicosities. Staining was blocked by preincubation of affiSB12 with the c-src( +) peptide antigen. from CEFlc-src and CEF/c-src( +) cell lysates, respectively, with monoclonal antibody 327 (lanes 1, 5), preimmune rabbit serum (lanes 2, 6), SB12 antiserum (lanes 3, 7), and SB 12 antiserum that was preadsorbed by the c-src(+) peptide antigen (lanes 4, 8). Immunoprecipitated src proteins were analyzed on a 7.5% SDS-polyacrylamide gel.
pp6@-sr@+) immunoreactivity was widely distributed in the rat brain. Schematic diagrams of half coronal sections through the rat brain demonstrating the distribution of pp60c-5rti+)-containing neurons are shown in Figure 3, while axonal profiles are shown in subsequent figures. pp60C-sr~+) was not detected in all neurons. Intensely immunoreactive neurons were visible in some areas, whereas moderately immunoreactive neurons were present in other brain regions. The following is a description of the distribution of pp60e+'&+) immunoreactivity through the rostralcaudal extent of the rat brain.

Pans/medulla
The pattern of pp60C-sr@+) immunoreactivity in the inferior olive represents one of the best examples of axonal pp6@-s'~+) immunoreactivity in the brain. Dense immunoreactive puncta were observed throughout the inferior olivary complex (Fig. 4). Scattered neuronal cell bodies were also stained. However, due to the dense immunoreactive puncta present in this structure, it was difficult to discern these immunoreactive neurons.
Densely stained neurons and neuropil were observed in the ventral tegmental nucleus (Fig. 54) and in the gigantocellular reticular area (Fig. 5B). Both the cell bodies and dendritic processes of these large neurons were densely stained.
In the nucleus of the trapezoid body, intense immunoreactivity was observed in neuronal cell bodies and their processes (Fig. 5C). In the superior olivary complex (Fig. 50, a dense plexus of immunoreactive axons was visible in the medial superior olivary nucleus as well as in the lateral superior olivary nucleus. Neuronal cell bodies and their dendrites were also stained throughout the superior olivary nuclei. Densely labeled neurons were also visible in the lateral reticular nucleus (Fig.  3H).
Several of the cranial nerve nuclei exhibited dense pp60C-sr6+) immunoreactivity. In the oculomotor nucleus, immunoreactive perikarya and dendritic processes were visible (data not shown). Intensely labeled neuronal cell bodies were observed in the mesencephalic trigeminal nucleus (Fig. 64. The spinal trigeminal nucleus also contained many pp60c-srd+)-immunoreactive cells, as well as clusters of immunoreactive puncta (Fig. 69. Immunoreactive perikarya, dendrites, and axons Gere visible in the facial nucleus (data not shown). In the cochlear nuclei, pp60"-srti+)-immunoreactive neurons were differentially distributed, with the dorsal subdivision containing a greater number of labeled cells than the ventral subdivision (Fig. 3G).

Cerebellum
The cerebellum contained some of the most intensely labeled neurons in the brain.pp6@-@+I immunoreactivity was localized to the cell bddy and dendritic processes of Purkinje cell neurons as well as in the cell bodies of granule cells (Fig. 724). At high magnification, dense labeling of the dendritic processes of Purkinje cells was visible in the molecular layer of the cerebellar cortex (Fig. 7B). Proximal as well as distal segments of the Purkinje dendritic arbor were immunoreactive. All of the deep cerebellar nuclei contained substantial populations of immunoreactive neurons (Figs. 3G, 7C') in addition to immunoreactive axons, as indicated by the punctate nature of the labeling (Fig. 7C).

Mesencephalon
In the substantia nigra (Fig. &I), p~6@-~~6+) immunoreactivity was restricted to the cell bodies and dendrites of neurons in the pars reticulata and the pars lateralis with no significant staining of neurons in the substantia nigra pars compacta. The immunoreactive neurons in the substantia nigra pars reticulata had diverse morphologies. Some neurons were small and ovoid whereas others were large and fusiform. The neuropil in the substantia nigra pars reticulata was also labeled, indicative of immunoreactive dendrites and axons. In the substantia nigra pars lateralis, small clusters of immunoreactive neurons were visible. Dorsal to the substantia nigra, intense immunoreactivity was also observed in the large neurons of the-red nucleus (Fig.  8B).

Diencephalon
Lightly stained pp60 -cTrd+)-immunoreactive neurons were scattered throughout diencephalic structures, including portions of the mammillary body, lateral geniculate and reticular thalamic nuclei, zona incerta, and lateral hypothalamus (Fig. 3, C, D). Intensely immunoreactive neurons were rarely seen in these brain regions, except for the mammillary nucleus, where a moderate population of densely stained neurons was visible (data not shown).

Telencephalon
The major telencephalic region exhibiting intensely labeled pp60c-Srd+)-immunoreactive neurons was the cerebral cortex,  where a distinct laminar organization was evident. In most areas of the cerebral cortex, only pyramidal neurons in layer V were densely labeled. Pyramidal neurons in layer VI of cortex were only lightly labeled. pp60C-sr6+) immunoreactivity was localized within the cell bodies and dendritic processes of pyramidal neurons in layer V of the cerebral cortex ( Fig. 94. The apical and basilar dendritic processes of these immunoreactive neurons were also heavily labeled (Fig. 9B). One exception to this layerspecific localization pattern was observed in the insular cortex where pp60C-Sr6+) immunoreactivity was distributed throughout all layers (data not shown). No significant immunoreactivity was detected in the cerebral cortex ventral to the rhinal fissure (Fig. 3, A-E).
A substantial population of pp60c-Sr@+)-immunoreactive neurons was also evident in the ventral pallidum, where the number of labeled neurons increased in the caudal direction (Fig. 3, A,  B). In addition to perikaryal and dendritic localization of p~6oC-~~d+), a dense plexus of immunoreactive axons was also present in the ventral pallidum ( Fig. 9, C, D). A moderate level of p~60E-~'6+) immunoreactivity was visible in the pyramidal neurons in fields CAl-CA3 of the hippocampus (Fig. 10). The remainder of the telencephalic structures contained scattered, lightly immunoreactive neurons. These areas include caudateputamen, globus pallidus, septum, and the central and basolateral amygdaloid nuclei (Fig. 3, R, C).

Discussion
We have described the distribution of the neuron-specific form of the c-src gene product, pp60c+'d+), in adult rat brain using an affinity-purified anti-peptide antibody that exclusively recognizes this variant form of C-SIC protein. These results confirm the neuronal specificity of pp60 c srd+), which was previously indicated by indirect studies in cultured cells (Bmgge et al., 1985) neurological mutant mice (Brugge et al., 1987b), and neurochemical lesion studies (Walaas et al., 1988). Furthermore, this study represents the first report of pp60C-Sr6+) expression at the cellular level using immunocytochemistry.
The distribution of p~6oC-~~d+) was widespread yet restricted  in the rat brain. pp60C-srN+) was not detected in all neuronal cells, The analysis of a closely related protein, pp561ck, which is but rather appears to be differentially expressed in subpopula-structurally homologous to p~6@-~'cc+) has shed some light on tions of neurons throughout the rat brain. Thus, pp60c-Sr<+) does the potential functions of this class of membrane-bound tyrosine not appear to be restricted to specific classes of neurons that protein kinases. This T-cell-specific protein is structurally and share an obvious common feature, such as morphology, neu-functionally coupled to the CDWCD8 membrane proteins, which rotransmitter responsiveness, or neurotransmitter/neuropep-serve as receptors for the major histocompatibility antigens tide localization. (Veillette et al., 1988(Veillette et al., , 1989. By analogy, pp60=(+) may asso-At the subcellular level, the distribution of pp60c-srd+) was also ciate with and be regulated by a neuronal cell receptor protein. widespread. p~6@-~~6+) immunoreactivity was localized within However, the pattern of expression of p~6oC-*~d+) in the adult the cell body and dendritic processes of certain neurons. In brain does not appear to correlate with the expression of any several discrete areas of the brain, including the inferior olive, known receptor protein molecule. Furthermore, pp6O~+~cc+) has superior olive, and spinal trigeminal nucleus, pp6O~+~a+) im-been detected throughout the nerve cell: in the cell body, denmunoreactivity was also localized to axons and nerve terminals dritic processes, axons, and nerve terminals. In addition, memas demonstrated by the presence of immunoreactive puncta.
brane fractionation studies have indicated that p~6@-~~6+) is en-The localization of p~60=-~~&+) within these functionally impor-riched in synaptic vesicles in adult rat brain (Pang et al., 1988a, tant   In the molecular layer of the cerebellar cortex, dense labeling was observed in proximal as well as distal segments of the Purkinje cell dendritic arbor. Medial is to the left. Scale bar, 100 pm. B, High magnification of a Purkinje cell neuron. Note the dense label in both the cell soma and the dendritic processes of this neuron. Scale bar, 20 Wm. C, High magnification of the medial cerebellar nucleus. Dense pp6W+'d+) immunoreactivity was visible in neuronal cell bodies and dendritic processes in the deep cerebellar nuclei. Immunoreactive puncta were also observed in all of the deep cerebellar nuclei. Medial is to the right. Scale bar, 100 pm. CCL, Granule cell layer; ML, molecular layer; XL, Purkinje cell layer. Figure 8. pp60'-*re+) immunoreactivity in coronal sections through the substantia nigra (A) and the red nucleus (B). A, High magnification of the substantia nigra. Note the dense labeling of neuronal cell bodies and dendrites in the substantia nigra reticulata. The neuro-pi1 in the substantia nigra reticulata was also labeled, indicative of immunoreactive dendrites and axons. No significant staining of neurons was visible in the substantia nigra compacta. Medial is to the left. B, High magnification of the red nucleus. Note the intense immunoreactivity localized to the cell body and dendritic processes of these large neurons. Medial is to the right. Scale bars, 100 pm. cp, Cerebral peduncle, basal, SNC, substantia nigra compacta; SNR, substantia nigra reticulata.
Our results confirm and extend previous c-src( +) localization immunocytochemical distribution of pp60c-Sr@+) is similar to the studies. Recently, an in situ hybridization study showed that pattern observed in the in situ studies; however, the regional C-SK(+) mRNA was localized to neuronal cell bodies using a distribution of c-src(+) mRNA was slightly different from the specific oligonucleotide probe (Ross et al., 1988). In general, the pattern of expression of p~6@-~~4+) we report here. While the in  situ studies showed very high levels of probe binding in the hippocampus, we observed only moderate levels of p~60~+~4+) immunoreactivity in this structure (Fig. 10). Biochemical studies also indicate that high levels of c-src( +) protein are expressed in the hippocampus (Ross et al., 1988;Walaas et al., 1988). One possible explanation for this difference is that a moderate level of p~6oC-~~4+) may be expressed in most neurons of the hippocampus, which contains a high density of neurons. This would explain why the levels of pp60 = srN+) -detected at the cellular level using immunocytochemistry appear lower than in biochemical assays in which the amount of pp60e-srN+) is assayed relative to total protein within a particular brain region.
Another difference between our immunocytochemical findings and previous results is that we observed some of the highest levels of p~60~*~4+) immunoreactivity in specific populations of cells within the brain stem, while the in situ hybridization study (Ross et al., 1988) and a previous biochemical study (Walaas et al., 1988) revealed low levels of c-src( +) mRNA and protein, respectively, in the pons and medulla. In those areas of the brain stem where we observed intense immunoreactive puncta, including the inferior and superior olive and the spinal trigeminal nucleus, one would not expect to see a similar pattern by in situ hybridization since mRNA is largely restricted to the cell bodies. Furthermore, in most biochemical studies the pons and medulla were homogenized together, thereby possibly preventing the detection of high levels of pp60c-srd+) expression in particular subnuclei in the brain stem.
Another member of the src family of tyrosine protein kinases has been shown to be expressed at high levels in the brain. Sudol et al. (1988) demonstrated that the proto-oncogene c-yes is expressed at high levels in the cerebellum, specifically in Purkinje cell neurons (Sudol et al., 1989). The dendritic processes of Purkinje cell neurons showed the most intense staining for pp62=-yes, the c-yes protein, whereas the cell bodies were less densely labeled. Our results indicate that pp60C-srN+) is also expressed in Purkinje cells. Dense pp60CJr~+) immunoreactivity was localized to cell bodies, dendritic processes, and axons of Purkinje cells in the rat cerebellum (Fig. 7). It is interesting that 2 such closely related proto-oncogenes are co-localized in one particular neuronal cell type. Perhaps each of these tyrosinespecific protein kinases is involved in the regulation of distinct processes. Unlike the c-yes protein, which is expressed at high levels in many different tissues, including brain, liver, kidney, and gonads, pp6Oc-r"+) is expressed exclusively in neurons. Therefore, it is possible that pp60c-srd+) may serve a neuronspecific function, whereas pp62e-ps may regulate a more general cellular function. Studies aimed at defining the precise subcellular localization of pp60C-srN+) and pp62E-yes should yield further information regarding the physiological significance of these tyrosine kinases in the Purkinje cell neuron.
Our immunocytochemical results indicate that pp60C-sr6+) was present at high levels in nerve terminals located in the inferior olive, superior olive, and spinal trigeminal nucleus, whereas low levels of p~6@'-~'d+) immunoreactivity were detected in nerve terminals in several brain regions, including the substantia nigra. Other evidence for the localization of pp6@+6+) in nerve terminals comes from deafferentation studies in which a significant loss of the c-src( +) protein was detected in the substantia nigra following lesions that interrupted afferent fibers coming from the neostriatum (Walaas et al., 1988). Previously, it was shown that synaptic vesicle proteins are phosphorylated by endogenous tyrosine kinases, and that pp60C-Src is expressed in nerve terminals and crude synaptic vesicle fractions (Pang et al., 1988a). More recently, Pang et al. (1988b) demonstrated that pp60C-rr6+) is present at high levels in purified synaptic vesicles. Furthermore, p~6oC-~~d+) was the most abundant tyrosine-specific protein kinase in synaptic vesicles. Together these studies suggest that pp6@-$"+) may play a role in signal transduction at the nerve terminal.
In summary, the widespread distribution of p~6oC-~~d+) in diverse neuronal cell types suggests the existence of a previously uncharacterized functional organization within the brain. The localization of this neuron-specific tyrosine kinase in functionally important areas of the nerve cell, namely dendrites, axons, and nerve terminals, suggests that pp60C-Sr&+) may regulate pleiotropic functions in specific classes of neurons in the adult central nervous system. Studies aimed at examining the developmental expression of pp60C-Sr6+) will provide further clues as to the physiological importance of this proto-oncogene in neurons. It will also be important to compare the cellular and subcellular distribution of the multiple members of the src family oftyrosine kinases that are expressed in the brain to obtain clues to the specific functions of these tyrosine kinases in the central nervous system.